Alternatives are limited for therapy of brain cancer and the patient outcome, despite radiation and chemotherapy treatment, is often death. We have developed recombinant biological agents based on targeting potent catalytic toxins that can address some of the major problems in the field. The first is that less than 0.001 % of the injected dose of any recombinant biological agent ever reaches the tumor and so we are investigating a solution by genetically engineering fusion proteins that can be administered directly into brain tumors using intracranial infusion. The second is that cell surface markers that are most commonly targeted in glioblastoma multiform (GBM), the most common form of brain tumor, are expressed on some GBM cases, but not on others. The solution was to broaden the range of tumors recognized by incorporating two ligands that commonly react against overexpressed GBM markers. This bispecific anti-GBM agent was called DTAT13. Another major problem is that an extensive network of vascularization readily nourishes surviving tumor cells contributing to GBM reoccurrence. We addressed this by directly targeting the vasculature. One of the ligands on our agent was directed against the amino terminal fragment (ATF) of urokinase that targets the human urokinase receptor (uPAR) overexpressed on tumor neovasculature. Finally, agents incorporating catalytic toxins have side effects that limit their dosage and our data indicate that the design of our agent reduces its toxicity. All of the work on DTAT13 to date has been performed on a mouse flank tumor model. In aim one, we will study our agent in a model that more closely mirrors our clinical use of these immunotoxins, i.e., a nude rodent model of intracranial therapy. The model will be used to establish working doses and dose schedules and address the critical issue of tumor reoccurrence. In the second aim, we will use primates to determine the pharmacokinetics, distribution into non-target organs, and to study the dose limiting toxicities in a model that more closely approximates its use in humans.